Left Brain Right Brain

Andrew Wakefield, the man who brought us the MMR scare, is back in the news. In a relatively small way. Earlier this year, Mr. Wakefield was in the news in a big way when the BMJ called him out for research fraud. At the time, Mr. Wakefield was enjoying the hospitality of wealthy patrons in a Jamaican resort, under the auspices of a “vaccine safety conference“. Also attending the conference were a team from the National Whistleblowers Center (aside, I am getting more and more leery of groups that put “national” into their name). Why is this important? Because instead of responding to the BMJ article directly, Mr. Wakefield provided material to one of members of the Whistleblowers Center so he could reply. The letter, by David Lewis, attempts to address one of the many issues involved with Mr. Wakefield’s research. It does not do this well, to be frank.

One problem with discussing Andrew Wakefield’s work is that there is just so much wrong that one can easily lose the forest for the trees. The big picture gets lost going over count after count of dishonesty and unethical behavior.

When analyzing Mr. Lewis’ argument, there are again many problems to discuss. Once again, it is easy to lose the story for the details. Allow me to address a few.

First, the argument is a classic strawman. Mr. Lewis posits that it is reasonable for Mr. Wakefield to have reported nonspecific colitis in his 1998 Lancet article, therefore no intentional misinterpretation was committed. To bolster this argument, he provided the BMJ with the original scoring sheets used by the pathologist who reported on the samples. Writes Mr. Lewis:

As a research microbiologist involved with the collection and examination of colonic biopsy samples, I do not believe that Dr. Wakefield intentionally misinterpreted the grading sheets as evidence of “non-specific colitis.” Dhillon indicated “non-specific” in a box associated, in some cases, with other forms of colitis. In addition, if Anthony’s grading sheets are similar to ones he completed for the Lancet article, they suggest that he diagnosed “colitis” in a number of the children.

As though, even if accurate, this would exonerate Mr. Wakefield. As the BMJ point out, in not one, but two pieces by specialists (gastroenterologist and pathologist), these grading sheets do not support the claims made in the 1998 Lancet article.

If you don’t have time to read those short articles, here’s the title of one: “We came to an overwhelming and uniform opinion that these reports do not show colitis”. Pretty clear.

And, yet, one can already see the battle lines being drawn and the talking points distributed. Before, people ignored the facts and instead tried to frame this as a personal battle: “Brian Deer, journalist, vs. Andrew Wakefield, Doctor.” Just a quick tweak and we have “David Lewis, expert who knows colonic biopsy analysis vs. Brian Deer, mere journalist.” This is, of course, more convincing that “Opinion of soil scientist vs. a pathologist and a gastroenterologist.” And, of course, easier than tackling the facts.

Let’s consider some facts, then. The first sentence of the (now retracted) 1998 Lancet paper: “We investigated a consecutive series of children with chronic enterocolitis and regressive developmental disorder. ”

“Consecutive series” was, at best, misleading given the facts of how the children were referred to the Royal Free. A fact not addressed in any way by Mr. Lewis.

“Chronic enterocolitis”. Note, “enterocolitis”. Not “non-specific colitis”. Let’s break this down. “-itis” means inflammation. Colitis “refers to an inflammation of the colon and is often used to describe an inflammation of the large intestine (colon, caecum and rectum).”

All fine so far. But, “enterocolitis“? The word actually used in the first sentence of the Lancet? The word used in the condition Mr. Wakefield coined for the condition he proposed, “autistic enterocolitis“? Enterocolitis is “an inflammation of the colon and small intestine.” And small intestine.

You can check the scores on grading sheets in the table Mr. Deer provided on his website. These include links to the scanned sheets themselves. For example, the sheets for Child 1.

The sheets and the table are arranged in a specific order: top down. From the highest point reached in the colonoscopy on down to the rectum (the exit of the large intestine).

Why point this out? Because you will notice from the table that for two of the eleven children reported (child one and twelve), there are no data for the duodenum (start of the small intestine) or the ileum (as noted in the Lancet article). Those reports start at the Caecum (the start of the large intestine) or lower. For some of the other children (child four, seven and nine), there are data, and they are listed as “normal”. The Lancet reports them as having lymphoid nodular hyperplasia. Child two has some notation, but again is checked “normal” by Dr. Dhillon. Likewise child 5 and child 6.

Which begs the question: how can you diagnose “enterocolitis”, inflammation of the large and small intestine, without data from the small intestine, or when the data you do have is normal? For 8 out of the 11 children, by my count.

Answer: you can’t. Hence the switch from “enterocolitis” to “non-specific colitis” in Mr. Lewis’ letter to the BMJ. In that letter, the term “enterocolitis” only appears in his citation of a Brian Deer article.

So, back to the first sentence of the Lancet article and ask, what is accurate? Let’s include the points made by Mr. Deer that not all these children demonstrated regression, and there wasn’t a single developmental disorder. So, with thanks to a reader who pointed this out: “We investigated a consecutive series of children with chronic enterocolitis and regressive developmental disorder(s).

or,

“We investigated children with developmental disorder(s)”

Not really very exciting.

While I have fallen into the trap I described at the outset, losing the forest for the trees, allow me to discuss one last piece of the way people are attempting to frame the way Mr. Wakefield wrote his paper. Instead of being an active researcher, it seems he is considered to be merely the man who collected the data and typed the manuscript. Mr. Lewis wrote, “I do not believe that Dr. Wakefield intentionally misinterpreted the grading sheets as evidence of “non-specific colitis.” ”

As though the entire process of diagnosing children with enterocolitis involved looking at Mr. Dhillon’s grading sheets. Once again, what did the Lancet article say? Under “Histology”:

Formalin-fixed biopsy samples of ileum and colon were assessed and reported by a pathologist (SED). Five ileocolonic biopsy series from age-matched and site-matched controls whose reports showed histologically normal mucosa were obtained for comparison. All tissues were assessed by three other clinical and experimental pathologists (APD, AA, AJW).

AJW refers to Andrew J. Wakefield. So, did he just accept the diagnoses from others, as we are being asked to believe now, or did he assess tissues, as he wrote in the Lancet in 1998? Seems an important point to downplay, as Mr. Lewis is not attempting.

To bring this to a close, a press release has been issued in support of Andrew Wakefield. Mr. Lewis is quoted:

“There was no fraud committed by Dr. Wakefield. The crux of the matter in Wakefield’s case, so far as research fraud is concerned, is whether Wakefield fabricated the diagnosis of non-specific colitis for 11 of the 12 Lancet children as claimed in Table 1. Drs. Paul Dhillon’s and Andrew Anthony’s grading sheets clearly show that Wakefield did not fabricate the diagnoses of non-specific colitis reported in the Lancet article.”

Well, no. Nonspecific colitis is not the “crux of the matter” at all. Mr. Lewis is referred to Mr. Deer’s article, and the discussion of regressive autism and when symptoms appeared. Facts which were not reported accurately in the Lancet.

And there’s that whole first sentence not being accurate at all thing…let’s not even get started on the concluding sentence:

We have identified a chronic enterocolitis in children that may be related to neuropsychiatric dysfunction. In most cases, onset of symptoms was after measles, mumps, and rubella immunisation. Further investigations are needed to examine this syndrome and its possible relation to this vaccine.

Today the British Medical Journal (BMJ) released a four articles on Andrew Wakefield and his research efforts at the Royal Free Hospital. To sum up the four articles in a single sentence (from Fiona Godlee of the BMJ): “Previously unpublished histopathology grading sheets apparently completed by Amar Dhillon, the senior pathologist on the paper, remove any remaining credibility from the claim that the Royal Free doctors had discovered a new inflammatory bowel disease associated with MMR”

For those who may wish a bit of background, earlier this year, the BMJ had a series of articles on Mr. Wakefield’s research:

Secrets of the MMR scare: how the case against the MMR vaccine was fixed (Feature by Brian Deer)

Secrets of the MMR scare How the vaccine crisis was meant to make money (Feature by Brian Deer)

The Lancet’s two days to bury bad news (Feature by Brian Deer)

and

Wakefield’s article linking MMR vaccine and autism was fraudulent (editorial by Fiona Godlee of the BMJ)

These articles came after Mr. Deer was able to review the transcripts of Mr. Wakefield’s “Fitness to Practice” hearings held before the General Medical Council (GMC).

Mr. Wakefield’s work focused on gastrointestinal disease. He went so far as to claim to have found a new syndrome, which he coined autistic enterocolitis. 13 years later, there is still no proof such a condition actually exists. When evidence came forward that the discussion of the children in the Lancet did not match the clinical records, Mr. Wakefield claimed that ““Dr Dhillon’s diagnosis formed the basis for what was reported in the Lancet,” and “I played no part in the diagnostic process at all.” Unfortunately it has been difficult to verify these statements. For one, the slides made from the samples taken from the children in the Lancet study are apparently missing, so direct comparison of what was reported in the paper to what was actually found in the children was difficult. Also, Dr. Dhillon’s analyses were not available. Until now. The scoresheets used by Amar Dhillon, the pathologist coauthor on the 1998 Lancet article, were recently made available to the BMJ. These form the basis for the articles published today.

In his new BMJ article, Pathology reports solve “new bowel disease” riddle, Mr. Deer discusses the pathology reports. He compiled a table based on these reports (for example, the report from Child 1). If I can summarize Mr. Deer’s discussion: what is striking about these reports is that they are not striking. Dr. Dhillon used his own score method to rank the disease found in the specimens. In most cases, these were “normal” or “mild”.

Mr. Deer shared the reports with a number of professionals for comment. Including Dr. Henry Appleman:

“Most of this stuff is so close to normal that you’ve really got to question whether there is really anything there,” said Henry Appelman, professor of surgical pathology at the University of Michigan and a specialist in gastrointestinal disease. “These are the kind of things that we in our practise here would ignore completely.”

And Ingvar Bjarnason:

“Everyone thinks I am crazy even asking them,” said King’s gastroenterologist Bjarnason, after discussing the scorings with other specialists. “All but one of the children is normal in their eyes. There is no enteritis and no colitis, simple as that.”

While the reports were not striking, there is indication of inflammation. Mr. Deer argues that a key point was left out of the Lancet article, the presence of constipation in many of the children, which would have put the inflammation signs into better perspective for the readers:

No less controversial, the authors omitted from the paper the children’s principal gastroenterological problem. “Almost all” had “severe” constipation.(30) The GMC panel heard, for example, that after bowel preparation by nasogastric tube, the first patient, who had mild caecal cryptitis, endured two attempted ileocolonoscopies that failed because of faeces in the caecum, with “scope trauma” noted on the second.

This omission of constipation was no small matter. It went to the heart of how the paper would be read. Specialists told me that both mild inflammation and prominent lymphoid follicles may be expected to be associated with this sign.

(30) Murch S, Thomson M, Walker-Smith JA. Authors’ reply. Lancet 1998;351:908.

Mr. Deer, of course, is not a medical professional. But, we do not have to rely solely on Mr. Deer’s report. The BMJ has two companion pieces by medical professionals.

In Commentary: I see no convincing evidence of “enterocolitis,” “colitis,” or a “unique disease process”, Karel Geboes, Professor Doctor, Department of Pathology Univ Hospital KUleuven reports:

In general, the data are rather similar to the reports of the Royal Free hospital pathology service, which I reviewed for the BMJ last year.(7) Although minor abnormalities are noted in a minority of patients, I see no convincing evidence of “enterocolitis,” “colitis,” or a “unique disease process” being present in all patients. The Wakefield et al paper is obviously problematic and its wording does not reflect the data shown in the grading sheets.

(7) Deer B. Wakefield’s “autistic enterocolitis” under the microscope. BMJ 2010;340:c1127.

In Commentary: We came to an overwhelming and uniform opinion that these reports do not show colitis, Dr. Ingvar Bjarnason, professor of digestive diseases, consultant physician and gastroenterologist at King’s College, London, writes:

The hospital pathology service found the histopathology to be normal,(3) and, except in the case of the child mentioned above, the grading sheets also note normal findings. The fact that these scores were interpreted as abnormal raises, in my opinion, questions for the authors of Wakefield et al to answer, and particularly for the consultant histopathologists.

From the histological and endoscopic reports, there are no grounds to believe that any new inflammatory bowel disease may have been discovered, or any possible “unique disease process” observed, as reported by Wakefield et al. Nothing can be said about the aetiology of any minor irritations noted, and nothing can be inferred regarding treatment.

(3) Deer B. Wakefield’s “autistic enterocolitis” under the microscope. BMJ 2010;340:c1127.

In Institutional research misconduct, Failings over the MMR scare may need parliamentary inquiry, BMJ editor Fiona Godlee discusses how there is no sign that the Royal Free Hospital (University College London) has begun the expected inquiry into the misconduct which occurred there in the 1990’s:

It is now more than 18 months since the UK’s General Medical Council found Andrew Wakefield guilty of dishonesty and other serious professional misconduct(1) ; and it is nearly a year since the BMJ concluded that his now retracted Lancet paper linking the measles, mumps, and rubella (MMR) vaccine with autism and bowel disease was an “elaborate fraud.”(2) (3) At that time, January 2011, we called on Wakefield’s former employer, University College London (UCL), to establish an inquiry into the scandal. Ten months on, no inquiry has been announced.

Ironically, these data were made available to the BMJ with the intent of exhonerating Mr. Wakefield.

In 1997, Dhillon was asked to reassess intestinal biopsy specimens taken from children enrolled in Wakefield’s research after the hospital’s histopathology service, under consultant and fellow coauthor Susan Davies, reported most of the children’s biopsies to be normal. His 62 A4 grading sheets were sent to the BMJ by c, a self employed environmental microbiologist. Lewis says he was given them by Wakefield after they met at a vaccine safety conference in January. In his accompanying letter, Lewis concludes that a non-expert pathologist such as Wakefield could have thought they showed that the children had non-specific colitis.

Ms. Godlee is being kind, in my opinion, using the title the self-styled “vaccine safety” conference chose. This was the same conference in Jamaica which was discussed here at Left Brain/Right Brain recently.

It strikes me as rather odd that Mr. Wakefield did not provide these documents to the BMJ directly. Perhaps in a response to the articles published earlier this year. What advantage he gained in providing them through a proxy is unclear.

When he was found guilty of misconduct by the GMC, Mr. Wakefield vowed that he would not go away. So far, this appears true. However, his arguments are getting weaker as time goes on. In this case, they seem to have outright backfired. Writes Ms. Godlee:

The grading sheets are certainly interesting, but not for the reasons Lewis (or, it may be assumed, Wakefield, in giving them to him) intended. We sent them to two independent reviewers and supplied the data for comment to two further senior gastroenterologists. We also showed them to Brian Deer, the investigative journalist who over the past eight years has uncovered the secrets behind the MMR scare and who arguably knows more about this case than anyone apart from Wakefield. Our expert reviewers are in no doubt that Dhillon’s findings—like Davies’s before him—are almost all normal, or as near to normal that the changes they reported were likely to be physiological.(6) (7) In an accompanying feature article, Deer explains what they add to our knowledge of the Wakefield saga.(8)

(6) Geboes K. Commentary: I see no convincing evidence of “enterocolitis,” “colitis,” or a “unique disease process.” BMJ 2011;343:d6985.
(7) Bjarnason I. Commentary: We came to an overwhelming and uniform opinion that these reports do not show colitis. BMJ 2011;343:d6979.
(8) Deer B. Pathology reports solve “new bowel disease” riddle. BMJ 2011;343:d6823.